🥛 Time to trim your crypto bags? 🤔

GM. This is Milk Road, the daily newsletter that's the floaty keeping you above water in choppy markets.

Here’s what we’ve got for you today:

✍️ Our Head of Research sold half his Coinbase stock.
🎙️ The Milk Road Show: While Everyone Is Chasing AI Stocks, This Indicator Says Buy Bitcoin Now w/ Matt Crosby.
🍪 Standard Chartered just started covering Uniswap.

Outskill is hosting a hands-on bootcamp designed to make you a Claude expert in just 16 hours. Sign up now for $0.

Prices as of 2:00 p.m. ET. Powered by CoinGecko.

OUR HEAD OF RESEARCH SOLD HALF HIS COINBASE STOCK… NOW WHAT? 💸

Last week our Head of Research, Kyle, did something that made a couple people blink.

He sold half his Coinbase (COIN) position - a stock that's up more than 4x from where he bought it.

Not because he's turned bearish. He's still long-term bullish on Coinbase (and he plans to keep the other half)...

The issue was that his portfolio had started to look like one big bet wearing five different jerseys.

Kyle holds BTC, ETH, COIN, HOOD, and GLXY. When crypto rips, they all rip. When crypto bleeds, they bleed together.

Source: TradingView

COIN was the heaviest piece of that, so trimming it softened more of his crypto risk than anything else would have.

Now he's sitting on fresh cash, and he wants to point it somewhere that doesn't move in lockstep with crypto.

(Wherever it lands, it'll show up in Milk Road PRO, same as all his trades.)

Which raises a fun question: where should he put his fresh capital?

We asked Vincent, our lead AI analyst - and his answer starts with what he's not touching.

And that is, the whole AI IPO wave:

SpaceX went public June 11 and raised $75B, the biggest IPO in history (more than double Saudi Aramco's old $29.4B record).
Anthropic filed its S-1 (the paperwork you submit before going public) on June 1 at a $965B valuation, with revenue running near $47B a year, up from about $9B at the end of 2025.
OpenAI filed too, reportedly chasing a $1T+ debut.

This is the obvious "diversify into AI" trade…

And Vincent is buying none of it.

BECOME A CLAUDE EXPERT FOR FREE

Claude is quickly becoming one of the most important AI tools of 2026.

And knowing how to use it properly is becoming a serious career advantage.

Good news:

Outskill is hosting a hands-on bootcamp designed to make you an AI-expert in just 16 hours.

Here’s what you’ll learn:

Master Claude’s Chat, Cowork, and Code modes
Set up plug-ins to automate your work
Build apps and dashboards without writing code (vibe coding, babyyy)

This session is completely free but only for the next 72 hours.

OUR HEAD OF RESEARCH SOLD HALF HIS COINBASE STOCK… NOW WHAT? (P2) 💸

Vincent’s problem is that this new wave of IPOs are priced almost entirely on growth, and growth that fast is hard to trust.

Run-rate (taking one recent month of revenue and annualizing it) tells you how fast money is coming in, not whether it sticks around.

Amazon and Google aren't only Anthropic's cloud providers - they're also two of its biggest investors.

Amazon has committed up to $25B into Anthropic. Google, up to $40B.

Then Anthropic turns around and pledges to spend that money buying their computing power, $100B promised to Amazon, with a matching setup at Google.

So the cash flows out of the cloud provider as an investment → then circles right back as cloud revenue (a solid chunk of that revenue number is the same dollars doing a full lap).

Source: Oakworth

And cheap Chinese open-source models (free AI anyone can download and run) are undercutting the labs from the other side, eating into their margins.

If the headline IPOs are the wrong place to look, where do you go?

Vincent's map points to a 5-layer "AI cake" that NVIDIA's Jensen Huang drew up:

Energy: The power that runs everything above it.
Chips: The GPUs that power feeds.
Infrastructure: The data centers that house the chips.
Models: The AI that runs on top (Anthropic, OpenAI).
Applications: Where the money finally shows up.

The IPOs everyone's chasing sit toward the top (the models & applications) - and the top is capped by a floor that can't grow fast enough to feed it (Energy, Chips, Infrastructure).

For example: only 5% of the AI chips already shipped are actually switched on.

The other 95% sit idle, waiting on power and data centers that don't exist yet.

JPMorgan pulled satellite images of planned sites and found more than 60% of the capacity meant to be live by 2027 hasn't even broken ground yet.

So the companies Vincent likes sit underneath the IPOs:

The ones squeezing more compute out of every watt, and the ones solving the power bottleneck entirely.

If you want to get a closer look at the exact assets Vincent is currently hunting, and get notified the moment he buys them - try Milk Road PRO for a buck for 7 days, here!

BITE-SIZED COOKIES FOR THE ROAD 🍪

Wanna invest in SpaceX but not based in the US? Check out Bitget Stocks 2.0 where you can trade major US stocks.*

Saylor: "I got very, very famous for saying you do not sell your BTC to the plebs." That advice was for retail.

Woah! Standard Chartered (a bank with nearly $1T in assets) just started covering Uniswap. Their 2030 price target on UNI is $100 (currently ~$2.70).

Raoul Pal: The market is totally missing the agentic economy - and the proof is in crypto.

Is this one of crypto's most underrated real-adoption plays? XDC sits right where RWAs, cross-border payments, and institutional money are all colliding.*

*this is sponsored content.